New project, need advices from you ;)

Hi all !
So, I decided to start my project, but only the design for the moment. And i need your advices !
Let's start with my idea's and what I want.
As I said in another topic ( therehttp://www.x-simulator.de/forum/what-about-this-rotative-solution-t1985.html, Thanks OscarH for your answer), I want to make my own motion simulator for 2 applications.
First, and it's the most important : for simulating flight, especially combat flight ( that's hard i think !)
And Second : for car-sim

For the first, i think that i need almost 2 Rotationnal DoF and another Z axis translation DoF.
I need too that my shuttle can turn 180°, for putting down my head ! :lol:
The Z translation axis is for simulating small drops that could appears. But it's mostly for the car-sim.
Indeed, in most car-sim, i think that we miss the suspension effect. That an important effect for me.

So for answering all these i designed a small 3D Part in solidworks :

what do you think about ?

http://www.youtube.com/watch?v=F-zmhcokfp8


I plan to use electric actuator, i hope i can find them in my company.

thanks you for answering and help me .

And sorry for my english !

Hi!

Nice simulation and its always fun when someone tries something new. :thbup:

Im not sure it will work that good though. First thought is that you will hit the yellow and the red frame as soon as you start leaning sideways. I think its better to do for example as tronicgr and let the outer frame lean side to side and the inner front and back. You will also have to put the actuators for Z axis translation in both ends, otherwise you will have to make the red frame much wider.

I also think it will be hard to find electric actuators that are able to move your simulator in Z axis translation. They have to be both strong and fast.

Regards
Tommy

If you want 360 rotation you should build this : maxflight-motion-simulator-t1140.html

Wow, I watched the video of the 360° Simulator and it's really amazing ! But no Z axis translation ( only for the start when passengers gets in).
But it's really impressive ! it's look like the shuttle can make a lot of round ! Do you know how they do that ? How is the data transmission ?
i know they are some technologies but I'm not familiar with them.

Otherwise, I didn't understand what you wrote, Tommy, about the fram that could hit each other ? really sorry, i think it's my english.
For the Z actuator, i saw some that could be fine, one member of X sim use them. There : http://www.x-simulator.de/forum/erics-6dof-ride-simulator-ride-u-lator-t1825.html
I think they will be a lot expensive but... I want to make it fine

However what do you think for moving the yellow frame en the grey one. I planed to use linear electric actuator with a disc ( as you can see) but I think that we could make it directly with a motor, no ?

Thanks you for answers !

1) if you mount seat on grey frame you can have 360° rotation you would have to have yellow mutch more bigger
2) power tramsition as you have (disk and acutator) is really bad idea, for exapmle you never get same torq.. if you can use direct motor or motor with gearbox.
3)Z if you want elxar acutators i thing you will be shock about price ...
4) about maxflight sim : you can buy parts for power transmition (i cant remember exact name for this things but they use it in robots). Z motion isnt problem if you use counter weight. But it move with whole sim, and with this amout of movement it do wrong effect

Ok, I understood what you wanted to say about the yellow frame.
I forgot to put a 1meter part on my img to see how is the size of the frames.
I'm gonna make another pic & video with this and a fast-designed cockpit.

About the actuator, I agree with you, i think that linear actuator wasn't a good solution because of my school lessons ( We start from a rotation to go to a translation an then another rotation - too much loss)

So, I can start with a motor and a gearbox as you say.

For the exlar price, I know it's expensive. But I think it will be the last improvement if I work well on my project :
First starting with a static red frame and then if it's work well, I will add these actuator.

More over, I plan to add some X & Y Actuator but just for simulating impact. They will probably be pneumatic or hydrolic with a little course, about 200mm. Just for moving hard the shuttle and come back to initial position.

Thanks for answering !

pics & videos coming in next post.

Here is the video showing the cockpit.

The yellow frame is 2 meters long and about 1 meter wide.

http://www.youtube.com/watch?v=fCiRqRTA-Jg

It's quite big

Nice
But watch out! Do not let him, to cut your head off!
How do you get in?
The point of rotating arm distance (diameter) very small.
This construction needs a very strong actuator, with a to short road, i think.
It would be necessary to increase the diameter possibly in practice.

Poi

In fact, the cockpit will be full closed, so no risk for my head !
No idea how i'll be get in, it's just some idea.
But maybe Mambo is right. I didn't think about the screen system for now. And for sure, the yellow frame will have to be larger.

One question more,
I'm really not familiar with motion control. As some of you say, a motor with gearbox will be more efficient to move yellow and grey frame. But I did'nt found a good explanation of wich type of motor will be ok. I heard ( and read) a lot about wiper motor.
I understand that we need to control the position the motor. So we need feedback from it no ?
And what is H bridge for ?
If someone have a link


Thanks !

For position so far only potenciometer. From what you want build i can tell you bigger motor better motor ...
Type of motor it depends on your budget.
H-bridge is something like amplifier you send low voltage control signal and you get bigger voltage and current on output.
I think you are to way ahead. First you must know how mutch you want invest in sim , what is your resource (if you have acces to machines atd. ) and then you can choise which way is better for you.
So far what you want will be BIG and expensive (if you want use 3 leg acutators).

I have access to a workshop at my work. We have machines for cylindrical part and prismatic part too ( I dont know the words in english - In french it's named Tour and Fraiseuse).
I have also access to a little CNC that i have been progamming last year for engineering small parts.
For the base actuator, as i say, i know that they will be expensive. But I know that we have some old's in the company garbage. That's why I want some advice for choosing them.

So I think engineering will not be a problem. The cost of the actuator will be one for sure, but I hope finding some actuators. However, in the first time, I can make the simulator without these 3 actuators.
I think that I will invest about 3000€. That's my goal. And make it in about 1 year or 1 year and half. I have to put 300€ each month, thats possible I think but it will be hard I have to see all possibilities that I have to reduce price. But I really want to make it well.


About the motor, I really don't understand how we can control position without feedback. You said that we need only a potentiometer. But with this we can control only the speed no ? Or is this very accurate and only with speed we can control position ?

Sorry for my questions but I'm in a mechanics school not electric !

I searched on wikipedia . They speak about step motor ( I think that's what we find on wiper motor no ? )and servo motor.
How brake these motor ? when they turn fast, they have inertia no ? So how they stop to there accurate position ?

Thanks - It's noobs question I know !

Sorry for my questions but I'm in a mechanics school not electric Tell me about it
You need only position because you want be in that position fast as possible.
Stepper motor isnt good for what you want (control, position feedback and power/price ratio ) For example most stepper has 1,8 deg/step sou you can get 200 position from it. Driver which control them mostly get 1 signal to direction of rotation and one signal to make one stepp. Motor itself can hold position you dont need fedback (in cnc machines, with x-sim i think you still need pot. ). If you overload motor you lose stepp you lose force and so on .. Stepper motor has bigest torq. in 0 rpm and than it fall fast and it can have big RMP (it can but price is bigger than better AC servos)

DC and AC servos can be control by analog signal (for example +- 10V (stepper can be also control by analog but its something like itch behind ear by leg .. )) postion is known from potenciometer (in future maybe by optical encoder). DC are cheap AC are expensive but more powerfull. Both can be overloaded (allmost 10x for short period of time, good for accelaration and braking). Difference is in control (DC have H-bridge , AC have freq. converter). If you have lots of money AC motor are the way, if you dont DC (you can get DC 1kW motor about 120 $ ).
I think you should bild something like 301, because 360° wont do exatly what you
want. Try to study how sim try to delude your sensation of movement and accel. . What would do what you want is this the-most-powerfull-flight-simulator-t1118.html but i dont think you have enoght room

Hi,

Mambo, maybe you readed into my mind ? lol . Seriously the two simulator that you showed me were the two that bringed me to DIY Motion sim. I saw the desdemona only on a webpage without photo or video about 2years ago. I just saw some little sketch and I was thinking that's amazing. Now when I see the video I'm really impressed ! The rotative frame can give a lot of G's ! But I'm still curious of the motion algorythm, I think it's probably very complicated !
The 301 is pretty cool too !

Let's get back,
I'm still misunderstanding how we can control position of motor ? you said with a potentiometer, but how ? i understand that we could control speed by adjusting voltage en current but for position ?
Without feedback how do we know if we are on the right or the left ?
I know it's very noobs question ! but I searched a lot in the forum and didn't found any answers ! maybe it's too easy !
THanks !

Potenciometer is connect to frame. For example from x-sim you send position 128 (1-256 for example) and AMC will try to get to position 128 as fast as possible (128 is half of rotation of potenciometer). So x-sim send position of sim and control board will get there and control board know position of sim by potenciometer (like joystick). Data wont change like from 1-128 instantly so if change is faster motion is faster if its change of positin slower speed is slower.

Desdemona motion algorythm isnt so complicated. Main arm rotate by constant angular speed. Cabin could rotate in x,y,z and slide on main arm. When there is only grav. force cabin is in middle so you feel only small centrifugal force. When there is force in X and Y cabin rotate in X and Y to force vector point out and change rotation radius proportionnaly to value of force. Main problem is that you need constant RPM of main arm and how you change rotat. radius of cabin arm want rotate faster or slower so you need it brake it or speed up it. What i dont get it is why they have linear motion in Z , if they could move inside cabin in x,y,z to simulate small movements (vibration, turbulance, ... ) i would get it , but only in Z ?